1. Field of the Invention
The present invention pertains to methods for the direct synthesis of compounds, e.g., polymeric or oligomeric compounds, that possess a complementary structure to a desired template molecule, e.g., a compound having biological activity. The present invention further pertains to compounds, e.g., polymers or oligomers produced by such methods, and the use thereof, e.g., as therapeutics or diagnostics based on their complementary structure to a molecule having a known activity. The direct synthesis methods provided herein, which are an extension of the technique generally known as "molecular imprinting," provide a powerful means of producing a compound having a desired activity. While the technique should be applicable for the synthesis of a complementary binding molecule to any desired compound, the most significant application comprises direct drug synthesis. As discussed in detail infra, the subject invention is particularly useful for direct synthesis of agonists or antagonists for desired molecules, e.g., enzymes, hormones, receptors and other proteins; molecules that affect gene expression, molecules that affect the binding of biomolecules; e.g., cells or cell-like moieties to other ligands; and the synthesis of improved diagnostic agents.
2. Description of the Prior Art
In traditional drug screening methods, natural products are generally isolated, e.g., from plant, animal or microbial extracts and tested for biological activity. These methods generally entail complex purification and characterization procedures, and the eventual identification of a natural product having biological activity, e.g., an antimicrobial agent. These natural products are used in their native form, or more typically they are improved by the synthesis of synthetic analogs thereof. These synthetic analogs are then tested for biological activity and the most active compounds become the "drug leads." These compounds are then used to develop the next generation of synthetic analogs.
While these methods have resulted in useful drugs, both natural and synthetic variants, they are generally very inefficient. Typically, testing must be carried out in animals, or potentially in vitro if there is a suitable in vitro model to test activity. This is problematic as many assays, in particular animal testing, require large quantities of compound. This is disadvantageous as it limits the number of compounds which can be feasibly tested.
Also, such methods are inherently complex and unpredictable. Often it is difficult to predict and establish the structure/activity relationship among different compounds tested for activity. This is difficult to assess, especially if the tested compounds vary significantly in structure. This makes it difficult to determine the particular portion of the molecule that is significant for activity. Generally, only by screening large numbers of compounds is this able to be determined.
Also, such methods are prone to error. Often compounds that score positive in in vitro assays, and even animal models, are inactive in humans. Conversely, compounds which score negative in vitro may actually be active but score negative because of solubility problems which enable an otherwise active compound to cross the cell membrane in vivo.
Recently, in an effort to obviate some of the problems and inefficiencies of traditional drug screening and synthesis methods, random screening techniques have been developed to identify active compounds. In such methods, a library, which is simply a collection of different chemical or biological entities, is screened for one or more properties, e.g., binding to a particular ligand. Such libraries include, by way of example, compound libraries, peptide libraries, oligosaccharide libraries, and nucleic acid sequence libraries. Typically, the compounds in a particular library possess a related structure, origin and/or function.
A particular type of library used by many research groups involved in drug design is the "combinatorial library." This simply refers to a library in which the individual members comprise systematic or random combinations of a limited set of basic elements. Randomization may be complete or partial. For example, some positions of the tested compounds may be fixed or varied systematically and others randomly varied. Typically, the members of a combinatorial library constitute oligomers or polymers, which vary based on the particular monomers, the connecting linkages, and/or the length of the oligomer or polymer. Ideally, the members of a combinatorial library are selected such that they can be screened for a particular activity or activities simultaneously. (See Fenniri, "Recent Advances at the Interface of Medicinal and Combinatorial Chemistry. Views on Methodologies for the Generation and Evaluation of Diversity and Application to Molecular Recognition and Catalysis," Curr. Med. Chem., 3:343-378 (1996), for a review of combinatorial library techniques.) One particular type of combinatorial library is the peptide library. These libraries may comprise peptides made by synthetic methods or by microbial synthesis. In particular, the use of phage or bacterial libraries wherein a phage particulate or bacterium expresses a desired peptide on its surface (by operable linkage of the corresponding DNA to a sequence that encodes a surface protein) are well known. These libraries are advantageous because peptides comprise structures that mimic many biological molecules, i.e., proteins. It is possible by synthetic or biological techniques to generate a large array of different peptides of a particular size and sequence, which are thereupon screened for a particular desired property. Microbial surface display libraries are advantageous in that large numbers of different peptides may be obtained in large quantities relatively efficiently. (See G. P. Smith and V. A. Petrenko, "Phage Display," Chem. Rev., 97:391-410 (1997), for a review on phase display libraries.) However, these methods also suffer significant disadvantages. In particular, peptides are often costly to synthesize, may be unstable (e.g., in the presence of proteases), and often are unable to cross cellular membranes. Therefore, other molecules, i.e., small organic molecules, still are preferred drug candidates.
Such compounds can also be screened by library screening methods. However, small molecules often are not trivial to synthesize in quantities necessary for screening. This disadvantage has somewhat been alleviated by recent methods which have downsized targets to the molecular level, and the automation of screens which have reduced the amount of compound necessary for assay to small amounts. These enhancements have enabled the utilization of combinatorial chemistry libraries instead of traditional chemical compound libraries. Combinatorial chemistry permits the rapid, relatively inexpensive synthesis of large numbers of compounds in small quantities suitable for automated assays directed at molecular targets. Numerous research groups and companies have reported the design of combinatorial chemistry libraries which exhibit a significant range of structural diversity. (See, e.g., P. M. Doyle, "Combinatorial Chemistry in the discovery and development of drugs," J. Chem. Tech. Biotech., 64(4):317-324 (1995); E. M. Gordon, "Libraries of non-polymeric organic molecules," Curr. Opin. Biotech., 6(6):624-637 (1995)). However, such screening processes still are often ineffective.
Thus, based on the foregoing, methods that provide for the direct synthesis of compounds having a desired activity, e.g., a desired biological activity would be highly desirable. Moreover, compounds generated by such methods would be extremely desirable because of their potential application as drugs and diagnostic agents.